R Notebook

library(Seurat)
library(dplyr)
library(tibble)
library(cowplot)
library(DT)

Data loading and processing

vijay <- readRDS('output/seurat_objects/vijay/vijay.rds')

Metadata plots

U/L branch marker genes

Plot the U and L branch marker genes.

#DE genes between T1T2T3 and T4T5 in the 10x-180831 data.
markers_u_l <- read.table('output/markergenes/180831/markers_10x-180831_upperbranch_lowerbranch_negbinom', sep='\t', header=T)

cannot open file 'output/markergenes/180831/markers_10x-180831_upperbranch_lowerbranch_negbinom': No such file or directoryError in file(file, "rt") : cannot open the connection

How do these genes look in the 180831 data?

And how are they expressed in the Vijay data?

Some U branch markers are not really expressed (SCD, PLIN4, GDP1, RBP4). Interestingly most U branch markers are also expressed in I1 and I3 and E1-E3.

Seurat integration with 10x-180831 data

Structural cells overlap with clusters P4, P5, P6, P2, P7. Metabolic cells overlap partly with P2 and P7 and also I3.

Predict cell types from 10x-180831 data

predictions2 <- TransferData(anchorset = anchors, refdata = data_180831$branch, dims=1:20, weight.reduction='pca', slot='pca')

Running PCA on query dataset

Scores

Assign the labels if threshold is above 0.7 (same threshold as the wolfrum label transfer).

for (col in c('pca_project', 'pca', 'cca.')){
  for (t in c(0.5, 0.7, 0.9, 0.95, 0.99)){
    preds <- assign_labels(col, t)
    #åvijay <- AddMetaData(vijay, preds, col.name=paste(col, 'predicted_label', t, sep='.'))
  }
}

Error in `[.data.frame`(vijay@meta.data, , c(paste(colname, ".prediction.score.max",  : 
  undefined columns selected

Threshold for prediction = 0.5

Threshold for prediction = 0.7

Threshold for prediction = 0.9

I3 = adipose tissue macrophages (Vijay paper). Perhaps no metabolic preadipocytes because of maturation?